1996-2000 Simple ECM fix to stop random dying

http://forums.cnet.com/7723-7811_102-248386/jeep-grand-cherokee-stalling-fixed/

This is probably already on here somewhere but I have never seen it and I could not find it when I searched. Anyway, it's worth posting again.
It's a band aid fix but it will get you back on the road. It just worked on my 1996 XJ.
My jeep was dead and the key-turn run codes said 11 (CPS). I replaced the CPS and it still didn't run. I spent all wknd trying to figure it out till I saw this website. Loosened 2 screws on the PCM and it's back to running normal again.

*scratches head* reading the article you posted makes no sense, what screws did you loosen on your ecu? are you talking about the screws that mount it to the chassis? or the ecu case screws? I dont understand because neither of those screws come close to the ECU board.

On my ECU, there are tiny torx screws. I loosened the two directly under the 3 electrical plug connectors.

BAM.., it no worky.., BAM.., it no worky.., BAM.., it worky.

Seemed to be easy enough to follow along, although a better choice of verbage describing sounds would've not distracted my comprehension, i.e., bad BAMs versus good BAMs.

Makes me glad I have a RENIX ECU, (ECM), (PCM), lol, LOL, lol, as the newer model he described seems to have a deal where you put the cover back on, then the PCM, or visa-versa in an earlier statement, blah, blah.

That c/net forum blurb was a hard read.., but the gist of the idea I got from it is that the screws holding the cover of the PCM on.., were too long, and touched some aspect of either a component, wire, or bread-board trace line causing for some 'short', or resistance throwing off the brain's ability to think straight, and thereby corrected by using shorter hold-down screws. Interesting.., and the fact that he may have gotten 'even' after the dealer run around he went through--well.., especially if he got his money back, i.e., the bottom line.

If anyone likes fiction, of the horror vein.., youtube: SPIDER DOG. Some will find it funny, i.e., get it.., and others may not, but anyways; Happy Chanukkah, Merry Christmas, and Happy New Year.

sidewaysstarion said:

I dont understand because neither of those screws come close to the ECU board.

Click to expand...

I've heard of a similar issue on the ZJ PCMs. But I don't recall what the issue with the screws is. I think the real problem lies in failed solder joints on the board where the sockets connect.

Talyn said:

I've heard of a similar issue on the ZJ PCMs. But I don't recall what the issue with the screws is. I think the real problem lies in failed solder joints on the board where the sockets connect.

Click to expand...

Well the ZJ's received two types of JTECs, both with the same part numbers but one made by Motorola and one made by Chrysler Huntsville electronics. Reading the article where he talks about the potting compound being hard to see through makes me think he was dealing with a huntsville jtec. The XJ chassis never was given huntsville ECU's that I've seen, I've only seen motorola ecus in XJ chassis. below you can see the difference in the two ECUs.



This is what the inside of a motorola Jtec looks like, as you can see the case screws are not near the board at all and there is no potting compound.


But I still have no idea what screws you were messing with or how they managed to fix your jeep.

Are the two torx screws you loosened those shown, through the bread-board, in the photo, (Thankfully provided by Talyn), left side, middle?

Alaskan89XJ said:

Are the two torx screws you loosened those shown, through the bread-board, in the photo, (Thankfully provided by SidewaysStarion), left side, middle?

Click to expand...

:wierd:

I have read some PCMs have a problem with the header that connects to the PCM's 5 volt power supply on a separate board from the mainboard.

Moving the header in one direction or another sometimes completes the cracks in the solder that's causing a problem with 5 volts power to the sensors.

Some people have tighten the header down using longer screws or wedged something to create tighter contact.

This is not a FIX. The PCM needs to be dismantled and the header re-soldered, if this is actually the problem. It could be something else needs to be repaired, but the cracks in the solder for the header is a known problem with some PCMs.

Muddzz said:

I have read some PCMs have a problem with the header that connects to the PCM's 5 volt power supply on a separate board from the mainboard.

Moving the header in one direction or another sometimes completes the cracks in the solder that's causing a problem with 5 volts power to the sensors.

Some people have tighten the header down using longer screws or wedged something to create tighter contact.

This is not a FIX. The PCM needs to be dismantled and the header re-soldered, if this is actually the problem. It could be something else needs to be repaired, but the cracks in the solder for the header is a known problem with some PCMs.

Click to expand...

If you look at the picture that I provided of the inside of the ecu...

sidewaysstarion said:

If you look at the picture that I provided of the inside of the ecu...

Click to expand...

All I can see is the outside of PCM and a V shaped circuit board and what appears might be the soldered header pins for the three C connectors.

Perhaps you could create a Youtube video, or add some more pics, so we can determine which parts of the PCM you're referring to and what may need to be repaired.

A couple of videos for the Grand Cherokee.

https://www.youtube.com/watch?v=uY62r-rtvls

https://www.youtube.com/watch?v=v8hdJwwx3_c

And another that uses spacers as a fix.

https://www.youtube.com/watch?v=hhbjUAzKT_E

Dodge jtec ecu repair

https://www.youtube.com/watch?v=gBLe8Vh5rNY

3/4 videos you linked were of Huntsville Jtec ecus, which are known to have issues and were never installed on the XJ chassis. the one video you linked with the XJ in it was kind of interesting, the reason I got my jeep so cheap is because it was having the exact same issue. (blowing 15 amp fuel pump fuse randomly). The previous owner replaced the PCM and the fuel pump and that didn't fix the issue. I found out if you wiggled the connectors sometime it would be weeks if not months before it would blow it again. One day I was backing up a driveway and the jeep cut out. I jumped out and went to throw another 15 amp fuse in it, as soon as I tried to install the fuse it blew, I unplugged the PCM and the fuel pump and it still blew the fuse. then I found the part of the harness that was rubbing on the fuel rail and shorting out from time to time taking that fuse out.

That's good to know some of the differences between the JTEC PCMs.

The Grand Cherokee grew out of the development of the XJ. The local JY use to have several Grand and XJ's that were undamaged, all had good parts, but the vehicles were around only for a few months. Then recycled in the cash for clunker program.

I carry a spare PCM and haven't ever had a reason to dismantle one.

The JTEC PCM used in later years XJ, TJ and ZJ are pretty much the same type of hardware design architecture. Uses one 16 bit and two 8 bit parallel processing microcomputers, capable of handling engines up to 10 cylinders @ 6,500 rpms.

Muddzz said:

The JTEC PCM used in later years XJ, TJ and ZJ are pretty much the same type of hardware design architecture. Uses one 16 bit and two 8 bit parallel processing microcomputers, capable of handling engines up to 10 cylinders @ 8,125 rpms.

Click to expand...

FiFY

Good, that's a maximum value specification?

The manual I have, the JTEC PCM description shows this type of hardware design (using three parallel processors) allows the main (16bit) processor to handle a higher volume of data processing (throughput) with reduced software complexity.

Manual states:

"10 cylinder engines running in excess of 6,500 RPMs "

Muddzz said:

Good, that's a maximum value specification?

The manual I have, the JTEC PCM description shows this type of hardware design (using three parallel processors) allows the main (16bit) processor to handle a higher volume of data processing (throughput) with reduced software complexity.

Manual states:

"10 cylinder engines running in excess of 6,500 RPMs "

Click to expand...

It's a software/hardware limitation anything over 8,125 rpms and the whole PCM shuts down, CCD messages stop, stops communicating over SCI, the whole 9 yards till the RPMS drop below that threshold. I figured this out after putting my trans into 1st at 80+ mph going down a hill... then the results were replicated on the bench.

So the XJ/ZJ PCM can work with the early Dodge Vipers?

benXJ said:

So the XJ/ZJ PCM can work with the early Dodge Vipers?

Click to expand...

No, no enough injector and ignition drivers to work on a viper, though you can reflash a viper computer and use it on a XJ/ZJ.

I have been having this random stall / cutout problem for a year now and nothing I do seems to fix it. Is there anything valid about this screw backout thing on a 1996 ODBII 1996 XJ sport ? looking at the one in my XJ and at the pictures I have no idea which screw to screw with ?